OPTICAL REMOTE SENSING TECHNIQUES FOR MONITORING OF INDUSTRIAL EMISSIONS

LAKI TISOPULOS, ANDREA POLIDORI, OLGA PIKELNAYA

South Coast Air Quality Management District, Diamond Bar, California

MOTIVATION

Adapted from Cuclis, 2012

ORS Refinery Measurement Surveys 1988 - 2008

• Optical Remote Sensing (ORS) technologies evolved significantly in the past decade

• Fully automated / continuous / no calibration required

• Ideally suited for long-term fenceline monitoring. Can characterize and quantify emissions

• Can be deployed from various mobile platforms for rapid leak detection, concentrations mapping and

emission flux measurements

• Measured VOC emissions can be higher (up to an order of magnitude) than those from emission

inventories

SCAQMD OPTICAL REMOTE SENSING MONITORING PROGRAM

• Demonstrate feasibility and effectiveness of

fenceline monitoring using optical remote sensing

• Improve LDAR program and reduce emissions

• Provide real-time alerts to downwind communities

• Measure actual facility-wide emissions

• Improve existing emission inventory estimates

2008 LP-DOAS for fenceline

monitoring. Contractor failed to fulfill obligations

2012 – 2014Two successful technology demonstration projects for

refineries

2015 ORS measurements campaign to study emissions from refineries,

small stationary sources and ships

2016-2018

Combined ORS and low-cost sensors deployments to study

impacts of HAPs on communities

2015 SCAQMD ORS PROJECTS

• Project 1: Quantify fugitive emissions

from large refineries

• Project 2: Quantify gaseous emissions

from small point sources

• Project 3: Quantify stack emissions from

marine vessels/ports

OPTICAL REMOTE SENSING METHODS

• Fluxsense Mobile Monitoring Laboratory equipped with SOF, FTIR, and DOAS

• Facility-wide emissions and real-time leak detection

• Top-down total emission estimates

• Surveyed all 6 major refineries in the South Coast Air Basin

• NPL DIAL facility • Bottom-up total

emission estimate

• Stationary OP-FTIR setup provided by Atmosfir can be used for long-term meas.

• Continuous 24/7 fenceline monitoring

Vertical Radial Plume Mapping

Differential Absorption LidarSolar Occultation Flux

• Kassay Field Services combined Stationary OP-FTIR and reverse plume modeling

Area Source Technique

PROJECT 1: QUANTIFY FUGITIVE EMISSIONS FROM LARGE REFINERIES

FluxSense SOF + FTIR + DOAS

Mobile measurements (daytime only)

5 week study at 6 refineries in the SCAB

Facility-wide emissions of methane, non-methane

VOCs, NO2, SO2, BTEX

National Physical Laboratory (NPL) DIAL

Stationary daytime and nighttime

measurements

1-week study at 1 refinery

Facility-wide emissions of non-methane

VOCs, BTEX

Ideal for field validation

Atmosfir Optics VRPM Using Open-path FTIR

Large installation, continuous (24/7)

measurements

5-week study at 1 refinery

Emissions of methane, non-methane VOCs

EPA OTM-10 method

Complements mobile and other short-term

observations

EMISSION MONITORING FROM REFINERY TANK FARM: EXPERIMENTAL SETUP

R1

R2

R3R4

R5DIAL

Atmosfir FTIR

SOF

Note: SOF track and DIAL lines of sight are approximate (for illustration only)

MONITORING OF A TANK LEAK EVENT

R1

R2

R3R4

R5DIAL

Atmosfir FTIR

SOF

Note: SOF position is approximate (for illustration only)

• October 5, 2015

11:30am-4:30pm

• Emissions from a tank

were observed by

all three ORS

technologies

• Fenceline

concentrations of

alkanes decreased

dramatically after

emissions stopped

EMISSIONS OF ALKANES FROM A LEAKING TANK

Fluxsense:

337+/_101 kg/h

NPL:

279+/_28 kg/h

DETECTION OF ELEVATED ALKANES AT REFINERY FENCELINE

Wind shifts

resulting in

elevated

levels at

fenceline

Leak repaired,

fenceline

levels declined

DISCOVERY OF UNDERGROUND LEAK FROM A CORRODED PIPE

• September 30, 2015, at ~4:00pm

• Fluxsense discovered a leak from a corroded underground pipe

• Discovery was made while driving inside the facility

• FLIR images/videos confirmed emissions from the ground

• Measured alkanes concentrations: ~70,000 ppb

• Average VOC emissions: 31 kg/h

Alkane column

[mg/m2]

Distance [m]

PROJECT 2: QUANTIFY GASEOUS EMISSIONS FROM SMALL POINT SOURCES

FluxSense SOF + Extractive FTIR + DOAS

Mobile measurements (daytime only)

5 week study of ~100 small sources:

Oil wells

Intermediate oil treatment facilities

Gas stations

Other small sources

Methane and non-methane VOCs, BTEX

National Physical Laboratory (NPL) Differential Absorption Lidar (DIAL)

Stationary daytime and nighttime

measurements

1 week study at selected sources

Methane and non-methane VOCs

Ideal for field validation

Kassay Field Services Open-path FTIR + reverse plume

modeling

Stationary daytime and nighttime

measurements

5 week study at ~50 small sources

Methane and non-methane VOCs, BTEX

OP-FTIR using EPA TO-16 method

• Good agreement between ORS techniques during co-

located measurements

• FTIR not able to capture the entire plume, but useful for

long-term trends

• FluxSense performed 24 mobile surveys

• Elevated NMHC emissions detected during all

monitoring days

EMISSIONS FROM A SMALL OIL TREATMENT FACILITY

October 09, 2015

Preliminary data

• Will insert FLIR video

DIAL visualization of VOC emissionsFLIR video

Well

Tank

Tree

EMISSIONS FROM A SMALL OIL TREATMENT FACILITY

• Storage tank is most likely the main source of emissions from the facility

PROJECT 3: QUANTIFY STACK EMISSIONS FROM MARINE VESSELS

FluxSense Mini-SOF, DOAS and ”traditional” methods

Measurements of individual ships

4 week study at Port of Los Angeles and Port

of Long Beach

Measurements performed

on-shore at fixed locations within POLA

and POLB

off-shore from R/V Yellowtail provided by

Southern California Marine Institute

“Real world” emissions (g/s) of SO2 and NO2

and “actual” emission factors (g/Kg fuel burnt)

of SO2, NOx and particulates from individual

ships

692 ships sampled during the study

Fixed measurement sites

Sample GPS track of

R/V Yellowfin

EMISSIONS FROM 692 SHIPS SAMPLED IN POLA AND POLB

IMO limit

NOx PM BC

Sulfur fuel content

Preliminary data Preliminary data Preliminary data

Preliminary data

AIRBORNE OPTICAL REMOTE SENSING MEASUREMENTS

Piper Archer Aircraft

MAX-DOAS Telescope

looking out of pilot’s window

MAX-DOAS Spectrometer on

the back seat

NO2 Column

Sunday, November 08, 2015

Preliminary data

AIRBORNE OPTICAL REMOTE SENSING MEASUREMENTS

Piper Archer Aircraft

MAX-DOAS Telescope

looking out of pilot’s window

MAX-DOAS Spectrometer on

the back seat

NO2 Flux

Sunday, November 08, 2015

Preliminary data

UPCOMING PROJECT:COMMUNITY-SCALE AIR TOXICS AMBIENT MONITORING

• Comprehensive 3-year study aiming to

1. use of ORS methods to monitor HAP

emissions from refineries and to estimate

their annual VOC emissions

2. use of ORS methods and “low-cost” sensors

for assessing the impact of industrial HAP

emissions on surrounding communities.

• Mobile ORS – detailed understanding of

emissions and concentrations mapping

(quarterly surveys)

• Low-cost sensors network – long-term

monitoring of VOC and PM2.5 around

fenceline and inside the community

Industrial Site

Adjacent Community

Low-cost VOC and

PM2.5 sensors

SOF

CONCLUSIONS• ORS techniques can provide:

• Quick identification of potential leaks, offering substantial improvement of LDAR program or ISD systems

• Detailed characterization of areas that contribute the most to measured emissions

• Real or near-real time emission measurements

• Improved emission inventories

• ORS methods are suitable for monitoring of emissions from large facilities as well as small sources

• Mobile ORS methods are effective way to screen large number of small sources quickly

• Good agreement between different ORS techniques during co-located measurements

• Strengths and weaknesses of each technology:

• SOF: mobile measurements are ideal for routine surveys inside and outside facilities

• DIAL: very precise and accurate, but not suited for long-term monitoring

• OP-FTIR: can provide useful information on long-term variability of emissions and record fenceline

concentrations of pollutants

ACKNOWLEDGEMENTS

• ORS contractors

• Johan Mellqvist, Jerker Samuelsson, Marianne Ericsson

FluxSense Inc., San Diego, CA

• Rod Robinson, Fabrizio Innocenti, Andrew Finlayson

National Physical Laboratory, Hampton Rd, Eddington, United Kingdom

• Steve Perry

Kassay Field Services, Mohrsville, PA

• Ram HashmonayAtmosfir Optics Ltd., Ein Iron, Israel

• Tesoro Carson refinery environmental staff for assisting with measurements inside the refinery tank farm

EXTRA SLIDES

METHODS: SOLAR OCCULTATION FLUX (SOF)

• Mobile measurements to record

total mass of molecules along

path traveled

• Total mass and wind data used to

calculate flux emissions (kg/s)

• Daylight measurements only

• Also used identify hot-spot areas

• Accurate wind data obtained

using SCAQMD’s LIDAR

• Vertical scans enable plume mapping

and flux calculation

• Combine integrated concentration with

simple wind field to obtain flux

• Can measure away from source

• Accurate wind data obtained using

SCAQMD’s LIDAR

METHODS: DIFFERENTIAL ABSORPTION LIDAR (DIAL)

• OP-FITR system is

positioned downwind from

the source

• Multiple retroreflectors

strategically placed to

cover outflow from the

source

• VRPM combines path-

averaged concentrations

from OP-FITR

measurements with wind

speed and direction to

calculate emission fluxesOP-FTIR

Emission

source(s)

reflectors

METHODS: VERTICAL RADIAL PLUME MAPPING (VRPM)

• Single light path OP-FITR

system is positioned

downwind from the source

• Retroreflector is placed so

emission plume crosses the

light path

• Path-averaged

concentrations from OP-FITR

measurements along with

wind speed and direction

are used to model emission

fluxes

• Reverse plume modeling

software is based on

Aermod

METHODS: AREA SOURCE TECHNIQUE